• Home
• What We Do
• Our Students
• Q&A with Kato Hussein Wabbi

Q&A with Kato Hussein Wabbi

What project are you working on?

I am developing a frugal Edge AI-based predictive maintenance system for critical medical equipment in low-resource healthcare settings and underserved communities. The system enables real-time, on-device fault and anomaly detection for early intervention, preventing equipment failures from escalating to breakdowns, reducing downtime, and improving care delivery. A key application is Virtual Buffering for vaccine refrigerators, a software approach that uses machine learning to replicate the thermal buffering of probes like glycol vials. The model learns the relationship between air and buffered temperature, estimating true vaccine-equivalent temperature in real time with standard air sensors. By removing the need for physical buffers, extra sensors, and repeated calibration, this innovation can cut sensor and maintenance costs by up to 70% while maintaining WHO and CDC compliance, making it scalable for low-resource settings.

What problem(s) are you solving?

In many low-resource hospitals, up to 40% of medical equipment is nonfunctional at any given time, often because equipment failures go undetected until breakdowns occur. This leads to downtime, which can last from days to months, leaving essential services unavailable. Maintenance in these settings is predominantly reactive, costly, power-intensive, and ineffective. As a result, many devices are abandoned, creating equipment graveyards, even when breakdowns are preventable. Edge Artificial Intelligence (Edge AI) offers a promising solution to preventing equipment breakdown by enabling real-time, on-device monitoring and early fault detection, without reliance on the internet or high power consumption. Detecting issues before total breakdown extends device lifespan, minimizes service interruptions, and improves patient outcomes. 

What brought you to this research?

I have witnessed firsthand that when you visit a hospital in a low- and middle-income country, one thing that stands out is the number of broken or faulty medical devices piling up in equipment graveyards or hospital corners. These devices are vital for healthcare delivery yet remain unused. When my supervisors introduced the idea of developing a predictive maintenance system for medical equipment, it immediately resonated with me because it addressed a challenge I had seen throughout my experience. Their guidance helped refine my focus, enabling me to channel my background in biomedical engineering into practical, frugal, and impactful solutions.